1. Field of the Invention
The invention relates to a method for determining electromagnetic waves originating from the interior of a melt, in particular a metal melt, especially in the visible light range and the adjacent UV range and infrared range, wherein a gas-filled hollow space is formed within the melt by blowing in gas and electromagnetic waves emitting from the melt are observed through the blown-in gas and evaluated by feeding the electromagnetic waves via an optical system to a detector for determining the temperature and/or chemical composition, as well as an arrangement for carrying out the method.
2. Related Background Art
In the production of steel in a converter or any other metallurgical reactor by refining pig iron or treating other melts in such a metallurgical vessel, it has always been endeavoured to have available as continuously and quickly as possible the temperature values of the melt and/or an analysis of the melt during the active treatment process in order to be able to keep the treatment process as short as possible and to get as near as possible to the target analysis sought. Rapidity is required, in particular, because the chemical reactions proceed at high speeds, involving the danger of being no longer able to interfere with the refining process or treatment process in due time. The extremely rough operating conditions prevailing in such plants do not meet with these objects. In the production of steel in a metallurgical reactor (converter, electric furnace, etc.), in the secondary metallurgical treatment of steel melts or in respect of any other non-ferrous metal melts (e.g., Cu, Ni, Al) it is, furthermore, advantageous to know the temperature and/or analysis of the melt after each treatment stage.
To solve these problems, attempts have, for instance, been made to get hints as to the correct point of time for terminating a refining process from the spectral analysis of the converter flame or from its absorptive effect relative to monochromatic light of a defined wave length. However, the strongly varying blowing conditions and the foaming slag on the melt bath as well as the high content of dust contained in the offgas do not allow for sufficiently precise conclusions on the temperature of the bath and the analysis of the melt.
Furthermore, it had been proposed for temperature measuring (DE-B-14 08 873) to insert into the refractory lining of the converter encapsulated thermocouples, which project into the converter interior and in the operating position of the converter come to lie below the meniscus of the melt to be refined. However, the durability of such thermocouples was insufficient; in addition, measuring results have been adversely affected by the necessarily strong cooling of the measuring device.
Furthermore, it is known to determine the temperature of a melt at a predetermined point of time by means of lances submerged in the melt. That method is disadvantageous if applied to the production of steel in a converter, because to that end the converter must be tilted and set right again, which involves a temperature loss of the steel bath of up to 40.degree. C. In addition, that method is time-consuming, because at first the blowing lance must be extended prior to tilting the converter and the converter must be set right again after having carried out a measurement, and it is only afterwards that the blowing lance--if necessary--can be retracted and blowing can be continued. Further drawbacks include that the measuring point within the melt may be chosen only arbitrarily, thus being hardly reproducible. Also the depth of immersion of the probe cannot be exactly determined, thus being hardly reproducible, either.
The determination of a chemical analysis of the melt is even substantially more complicated. To this end, it is known to take samples by means of lances submerged in the melt. When producing steel in a converter this involves disadvantages since the taking of such samples likewise requires much time--the converter likewise having to be tilted (except with perpendicular sublance measurements)--and the samples must be sent to the laboratory.
When producing steel in a converter it is known to carry out a quick carbon analysis by measuring the arrest point of the temperature and the C content. Thereby, it is however only feasible to acquire the C equivalent such that some of the accompanying elements present in the melt have to be taken into account when calculating the actual carbon content.
Furthermore, it is known to carry out carbon and oxygen activity analyses and to take samples and temperature measurements in a converter by means of sub lances. This is, however, disadvantageous inasmuch as the sublance means themselves (and also the samples) are very expensive, prone to extremely high wear and applicable only with liquid slags towards the end of a blowing process.
From EP-B-0 162 949 a method for observing the formation of slag in a steel blowing converter, using the light radiation emitted from the slag surface within the converter space is known. There, the light is photoelectrically converted into signals and processed, variations of the signals being taken as criteria of foamed slag formation. The receptors inserted in the side wall of the converter are located above the slag/melt bath and are not suitable for measuring the melt bath temperature and the melt composition.
From U.S. Pat. No. 4,830,601 a method and an arrangement for the spectral-analytical evaluation of the light emitted from the central portion of a burner flame is known. There, the supply of fuel and combustion air is surveyed by way of the light spectrum. Emitted light is transmitted to an electronic evaluation device via fiber-optic conductors, the supply of combustion air and fuel being controlled as a function of the gas analysis made.
A similar arrangement for temperature measurement in a process for producing reducing gas in a high-temperature reactor at an elevated operating pressure is to be taken from DE-A-40 25 909.
From EP-A-0 214 483 it is known to verify the chemical composition of iron by blowing oxygen or an oxygen-containing gas from top onto the surface of molten iron, wherein beams originating from the melt surface are detected in a spectrometer with a view to determining the chemical composition of the iron.
From U.S. Pat. No. 4,619,533 and EP-A-0 362 577 methods of the initially defined kind are known, wherein in the first case radiation originating from the metal melt is conducted to a detector via a fiber-optic waveguide. According to EP-A-0 362 577 laser light is focussed on the metal surface thus generating plasma. The plasma light emitted from the metal surface via a lens system and a fiber-optic waveguide is fed to a spectrometer for elementary analysis. The lens system comprises adjustable lenses. The lenses are adjusted in a manner that the ratio of the intensities of two iron lines, namely the intensity of an atomic line arid the intensity of a ionic line, is minimal.
In a method of the initially defined kind, i.e., when detecting electromagnetic waves originating from the interior of a melt, the blowing in of gas for the formation of a gas-filled hollow space advantageously is effected through a wall opening of a metallurgical vessel receiving the metal melt, said opening having to be located below the standard meniscus. In the region of transition of said opening of the metallurgical vessel towards the melt, i.e., in the marginal region of said opening, reflections of the electromagnetic waves emitting from the melt are caused even with a very small opening, leading to falsifications of the measured values. If a mushroom-shaped incrustation is formed of solidified melt as a result of the blown-in gas, the incrustation having the shape of a bead surrounding the marginal region of the opening about the total periphery and oriented in the direction towards the melt constitutes a disturbing factor despite its protective function for the opening, constantly varying in size and position, whereby radiation originating from the surface of the incrustation or from the region of transition of the incrustation to the melt will falsify the measuring result. It has been shown that a precise measurement can be carried out only if radiation originating exclusively from the melt surface is received and transmitted to the detector. Reflections from the marginal region of the opening or from the incrustation are strongly disturbing, i.e., bring about falsifications of the measured values, without this being recognizable by any other indications.